Logistic Regression Accuracy
0.6216
Linear SVC Accuracy
0.5980

Logistic Regression Classification Report

              precision    recall  f1-score   support

    Democrat       0.71      0.78      0.74        59
 Independent       0.69      0.68      0.68        56
  Republican       0.86      0.78      0.82        54

    accuracy                           0.75       169
   macro avg       0.75      0.75      0.75       169
weighted avg       0.75      0.75      0.75       169

Linear SVC Classification Report

              precision    recall  f1-score   support

    Democrat       0.69      0.76      0.73        59
 Independent       0.67      0.64      0.65        56
  Republican       0.84      0.78      0.81        54

    accuracy                           0.73       169
   macro avg       0.73      0.73      0.73       169
weighted avg       0.73      0.73      0.73       169

Dashboard Summary

  • This dashboard compares two classification models, Logistic Regression and Linear SVC, for predicting political affiliation.
  • Logistic Regression achieved the higher cross-validated accuracy (0.6216), outperforming the Linear SVC (0.5980).
  • Because Logistic Regression is both more accurate and interpretable, only its confusion matrix and feature importance plot are shown.
  • Feature importance highlights which survey responses most strongly influence political affiliation predictions.
  • Both models' classification reports provide precision, recall, and f1-score across all classes.
  • Overall, Logistic Regression offers the best balance of interpretability and predictive performance for this dataset.